[RESEND RESEND PATCH v2] mtd: nand_bbt: scan for next free bbt block if writing bbt fails

Xander Huff xander.huff at ni.com
Tue Aug 11 14:55:22 PDT 2015


From: Ben Shelton <ben.shelton at ni.com>

If erasing or writing the BBT fails, we should mark the current BBT
block as bad and use the BBT descriptor to scan for the next available
unused block in the BBT. We should only return a failure if there isn't
any space left.

Based on original code implemented by Jeff Westfahl
<jeff.westfahl at ni.com>.

Signed-off-by: Ben Shelton <ben.shelton at ni.com>
Reviewed-by: Jaeden Amero <jaeden.amero at ni.com>
Suggested-by: Jeff Westfahl <jeff.westfahl at ni.com>
Signed-off-by: Xander Huff <xander.huff at ni.com>
---
This v2 patch is in reply to comments from Brian Norris on 7/22/13. See
the following links for context:
http://lists.infradead.org/pipermail/linux-mtd/2013-July/047596.html
http://patchwork.ozlabs.org/patch/244324/
---
 drivers/mtd/nand/nand_base.c |  4 ++++
 drivers/mtd/nand/nand_bbt.c  | 34 ++++++++++++++++++++++++++++++++--
 include/linux/mtd/nand.h     |  7 +++++++
 3 files changed, 43 insertions(+), 2 deletions(-)

diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c
index ceb68ca..48299dc 100644
--- a/drivers/mtd/nand/nand_base.c
+++ b/drivers/mtd/nand/nand_base.c
@@ -2761,6 +2761,7 @@ int nand_erase_nand(struct mtd_info *mtd, struct erase_info *instr,
 		pr_debug("%s: device is write protected!\n",
 				__func__);
 		instr->state = MTD_ERASE_FAILED;
+		instr->priv = NAND_ERASE_WRITE_PROTECTED;
 		goto erase_exit;
 	}
 
@@ -2776,6 +2777,7 @@ int nand_erase_nand(struct mtd_info *mtd, struct erase_info *instr,
 			pr_warn("%s: attempt to erase a bad block at page 0x%08x\n",
 				    __func__, page);
 			instr->state = MTD_ERASE_FAILED;
+			instr->priv = NAND_ERASE_BAD_BLOCK;
 			goto erase_exit;
 		}
 
@@ -2802,6 +2804,7 @@ int nand_erase_nand(struct mtd_info *mtd, struct erase_info *instr,
 			pr_debug("%s: failed erase, page 0x%08x\n",
 					__func__, page);
 			instr->state = MTD_ERASE_FAILED;
+			instr->priv = NAND_ERASE_BLOCK_ERASE_FAILED;
 			instr->fail_addr =
 				((loff_t)page << chip->page_shift);
 			goto erase_exit;
@@ -2819,6 +2822,7 @@ int nand_erase_nand(struct mtd_info *mtd, struct erase_info *instr,
 		}
 	}
 	instr->state = MTD_ERASE_DONE;
+	instr->priv = NAND_ERASE_OK;
 
 erase_exit:
 
diff --git a/drivers/mtd/nand/nand_bbt.c b/drivers/mtd/nand/nand_bbt.c
index 63a1a36..09f9e62 100644
--- a/drivers/mtd/nand/nand_bbt.c
+++ b/drivers/mtd/nand/nand_bbt.c
@@ -662,6 +662,7 @@ static int write_bbt(struct mtd_info *mtd, uint8_t *buf,
 			page = td->pages[chip];
 			goto write;
 		}
+next:
 
 		/*
 		 * Automatic placement of the bad block table. Search direction
@@ -787,13 +788,42 @@ static int write_bbt(struct mtd_info *mtd, uint8_t *buf,
 		einfo.addr = to;
 		einfo.len = 1 << this->bbt_erase_shift;
 		res = nand_erase_nand(mtd, &einfo, 1);
-		if (res < 0)
+		if (res == -EIO && einfo.state == MTD_ERASE_FAILED
+		    && einfo.priv == NAND_ERASE_BLOCK_ERASE_FAILED) {
+			/* This block is bad. Mark it as such and see if
+			 * there's another block available in the BBT area. */
+			int block = page >>
+				(this->bbt_erase_shift - this->page_shift);
+			pr_info("nand_bbt: failed to erase block %d when writing BBT\n",
+				block);
+			bbt_mark_entry(this, block, BBT_BLOCK_WORN);
+
+			res = this->block_markbad(mtd, block);
+			if (res)
+				pr_warn("nand_bbt: error %d while marking block %d bad\n",
+					res, block);
+			goto next;
+		} else if (res < 0)
 			goto outerr;
 
 		res = scan_write_bbt(mtd, to, len, buf,
 				td->options & NAND_BBT_NO_OOB ? NULL :
 				&buf[len]);
-		if (res < 0)
+		if (res == -EIO) {
+			/* This block is bad. Mark it as such and see if
+			 * there's another block available in the BBT area. */
+			int block = page >>
+				(this->bbt_erase_shift - this->page_shift);
+			pr_info("nand_bbt: failed to erase block %d when writing BBT\n",
+				block);
+			bbt_mark_entry(this, block, BBT_BLOCK_WORN);
+
+			res = this->block_markbad(mtd, block);
+			if (res)
+				pr_warn("nand_bbt: error %d while marking block %d bad\n",
+					res, block);
+			goto next;
+		} else if (res < 0)
 			goto outerr;
 
 		pr_info("Bad block table written to 0x%012llx, version 0x%02X\n",
diff --git a/include/linux/mtd/nand.h b/include/linux/mtd/nand.h
index 272f429..86e11f6 100644
--- a/include/linux/mtd/nand.h
+++ b/include/linux/mtd/nand.h
@@ -1030,4 +1030,11 @@ struct nand_sdr_timings {
 
 /* get timing characteristics from ONFI timing mode. */
 const struct nand_sdr_timings *onfi_async_timing_mode_to_sdr_timings(int mode);
+
+/* reasons for erase failures */
+#define NAND_ERASE_OK			0
+#define NAND_ERASE_WRITE_PROTECTED	1
+#define NAND_ERASE_BAD_BLOCK		2
+#define NAND_ERASE_BLOCK_ERASE_FAILED	3
+
 #endif /* __LINUX_MTD_NAND_H */
-- 
1.9.1




More information about the linux-mtd mailing list